A method for comprehensively analyzing long-term time series of distributions of informative aquatic environmental parameters determined from satellite data has been proposed and tested for physically based zoning of vast marine areas. The method is based on an approach that combines statistical, correlation, and regression analysis, as well as principal component analysis and self-organizing classification of heterogeneous data. Using the results of processing and analyzing daily spatiotemporal distributions of physical parameters determined from satellite and model data, including sea surface temperature, sea ice concentration, sea level, and sea surface salinity with a retrospective depth of ~ 30–40 years, the characteristics of zones with anomalous dynamics of these marine environmental parameters have been identified and analyzed for waters near the Kamchatka Peninsula, in the Barents Sea, and in the Atlantic Antarctic. The research results have demonstrated the feasibility of the proposed method for solving problems related to detecting areas of seas and oceans characterized by interannual variability in the physical characteristics of the marine environment that is atypical for the considered region.
Bondur et al. (Mon,) studied this question.